Michael Aldrich BiographyHistory of ROCCAcknowledgementsFeedbackBack to Aldrich Archive Home

  Internet Online Shopping

  Inventor's Story

  Online Shopping in the 1980s


  E-commerce, E-Business and Online

  What is Videotex?

  Videotex New Generation Terminals

  Videotex Chips

  "Videotex: Key to the Wired City"
        USA Book Review

  "Videotex: Key to the Wired City"
        German Book Review

  I Hate Shopping

  Software Anecdotes

  Emerging Digital and Social Markets
Digital People Power Changes the World

  Online Shopping FAQ's


Software Anecdotes
Memories of Viewdata Plus - Johnny A.S.P. Walsh
Johnny A.S.P. Walsh worked at Redifon, Rediffusion and ROCC Computers between 1978 and 1994: Systems Analyst, Software Engineering Operations:  Sales Account  Manager : Product Manager: Project Manager: Consultant.

In late 1979, Martin Dines  and I worked in National Support for John McGregor-Temple. At that time, our role in the team was to investigate software problems and determine whether they were system software bugs or required other remedies. The team supported  Keycheck and 7E software for Seecheck, and the new A and B load software releases.

We were keen data prep application programmers, but we believed we were useful to the system software experts, assembler programmers and communication geeks because we could perhaps better communicate the needs of customers, QA the patch fixes and advise customers on workarounds.

Everyone was excited by the new B3 release software, with its potential for online applications using the new database functions , in addition to everything we already had for data prep. Before B release, you could write sophisticated applications in 7E and A load, if you squeezed the last drop of functionality from indexed sequential field value tables and used “command sequences” to maintain batches as if they were databases.

It was as application programmers, that Martin and I were called upon to write the user demonstrations for the launch of Viewdata Plus, being held at the Effingham Park Hotel along with the new Writaway handprint recognition tablet. It was some weeks away, and there were parallel streams of work in other areas, not least to advance the host modems from the Tupperware housing that Peter Champion and Jim Bethell had shown us. We decided to write applications to first show Viewdata Plus as a Prestel-like page information system, but then to show it off as an online database terminal working from a Rediffusion telly anywhere with a phone socket and a remote pad. Pre-release software was not an unknown species to us, and we had to muddle through with bugs and early performance problems. At the time, we also did not have functions to store the whole hexadecimal codeset as data, so we had to invent our own escape sequences and interpreter application. But on the bright side, we did have a portacabin at our disposal for our around the clock programming and debugging of dems, warmed by the Ampex disk drives, each the size of an American fridge, as used by the Keycheck support team that shared our humble abode. The page-based dem attempted to show off page and interactive graphics by reproducing the activity of carbon rods regulating the rate of fission in a nuclear power reactor.

Supporting Advanced Viewdata Plus
Mike Aldrich has described in his paper how Viewdata Plus took off. To keen eyed entrepreneurs and public servants, the lucrative novelty of interactive Viewdata Plus over one-way and page-based Viewdata had been recognised. Many commercial contracts involved non-disclosure agreements so our keener customers could be the first to offer smart interactive applications from  car showrooms, travel agents and homes using an advanced telly with a phone cable.

The challenge for us was to keep the system software robust and resilient, and to build bigger systems to support larger populations of busy Viewdata tellies. Bigger systems came in the form of ROCC 1850 and then the 2800 series with the C software release, later re-named Workstation Management  System[WMS]. These were still the days before distributed databases, and just prior to Winchester technology sealed disk drives.

Memories include:

  • Talbot Motors director hammering the Videotex keyboard with his fists to mimic what his staff might do if there was a problem

  • Relief when, late one night, I was able to reproduce a critical bug happening on Ford’s live systems. I had managed to mimic busy live systems with my program equivalent to a rampant ferret down a rabbit hole.

  • Trying to keep my balance on the thickest  pile carpet in Nissan Director offices while swearing oaths to my grandmother to fix bugs

  • Using the UAPT Viewdata credit history service to find any country court judgements on pretty much everybody I knew

  • Acting like a SWAT team and using choppers between Gatwick and Heathrow to fly off to fix urgent bugs

Mainframe Passthru
I started as a data prep analyst/programmer but advanced my knowledge by a combination of ongoing training programmes, being surrounded by geeks enthusiastic to share their knowledge, and plenty of reported incidents from a large and demanding customer base.

My first taste of IBM 3270 emulation came at BEREC (British Ever Ready Electrical Company) in Wolverhampton in around 1980. They had bought several R850s to use for data prep as well as replacing expensive IBM 3270 terminals. The 3270 functions needed close support, and Tony James would lug a datascope and listings to sites so Tony could fix 3270 bugs in situ (since we didn’t have our own mainframe at base).

I don’t recall who developed 3270 Passthru but once I saw how it worked and what we could do with it I was its biggest fan and programmed with it into the night. It’s common now to use Application Program  Interfaces (APIs) to interface different applications, but in the early 1980s I marvelled at how I could write smart robotic programs to talk to the mainframe as if a user was tapping keys on its own 3270 terminals. 3270 Passthru could extend the scope of mainframe applications through friendly, easy to write ROCC applications whether the user was using a ROCC VDU, a Viewdata TV, or a Writaway pad.

I wrote a development tool and diagnostic program, tested against the RHM mainframe service I could use at Crawley. My program ran on Viewdata or VDUs, and initially was mostt used for demonstrating to utterly surprised prospects how we could manipulate, enhance and extend their stable and complex COBOL applications without changing the mainframe at all. In particular, I recall goggle-eyed Data Processing Managers (DPMs ) at many a demonstration. It also meant I got out of the office a lot more on sales support rather than bug fixing, and ended up giving somewhat over enthusiastic courses on 3270 Passthru at our training school.

Viewdata and Passthru Sales
When I moved to Manchester and into Sales, I sold Viewdata Plus and data prep and comms systems in North East England and East Midlands. With Dave Bunting and the Manchester team, we installed Viewdata Plus at North East Regional Gas (NORGAS) in Leeds for their Executive Information System (EIS). This was a novel way of using interactive Viewdata Plus to turn huge ICL print report spools into easily digested and colourful charts and tables on the desks of directors and senior managers who wouldn’t otherwise touch a keyboard. The EIS appeared to seen as a rite of passage in the NORGAS management hierarchy. After showing how to turn data into Viewdata charts, NORGAS wrote their own programmes with conventions to show favourable data in green, and adverse data in red, no matter the signage of the data.

This was before networked personal computers with spreadsheets could be used to download mainframe data and be accessible to non-technical users.

Other EIS and Viewdata Plus projects were implemented by the Manchester team. The field engineering system at Stelrad in Hull transformed mainframe applications using 3270 Passthru so that heating engineers around the country could see their schedules and order their parts from TVs at home.

December 2010



Biography                  History of ROCC                  Acknowledgements                  Feedback                  Copyright

© Michael Aldrich 2011